1). Field of the Invention
The present invention relates generally to a method and an apparatus for plating onto a substrate. More specifically, the present invention relates to a method and apparatus for controlling thickness of plating over a width of a substrate. In particular, the present invention relates to a method and apparatus for counteracting non-uniform plating over a width of a face of a substrate.
2). Discussion of Related Art
Plating techniques are widely known in the art and are used for a wide variety of purposes. In the microelectronic industry, for example, electroplating and electroless plating techniques are used for plating layers on a face of a substrate or for forming individual structures on substrates, such as on semiconductor wafers during the manufacture of integrated circuits, on semiconductor dies, on printed circuit boards, and on various other active and non-active electrical components. For example, the manufacture of an integrated circuit on a wafer typically involves the formation of a pattern of metal lines on the wafer. The metal lines may be plated within trenches formed in a surface of the wafer. The metal lines may be formed utilizing an electroplating technique wherein a voltage is applied to a metal shorting layer in a base of the trenches. Previously, metal was plated which filled the trenches and covered the wafer, whereafter the metal would be planarized to leave a layer of metal lines in the trenches only. In order to do away with a planarization step, techniques were then developed wherein only the trenches were filled with metal without covering all of the wafer. Such a technique requires uniform plating over the entire width of the wafer so as to ensure uniform metal line thicknesses over the width of the wafer. However, for various reasons some of which will be discussed hereinbelow, uniform plating of a layer over a width of a wafer is often not easily accomplished.
In another example, individual contact structures known as controlled collapse chip connection (C4) bumps are formed on a wafer after an integrated circuit is formed in the wafer. The C4 bumps are used to mount the integrated circuit to a package substrate. The C4 bumps are formed by applying a voltage to individual bond pads on the wafer. In order to ensure C4 bumps of substantially equal height, as is required for mounting the integrated circuit to the package substrate, uniform plating over the width of the wafer is required. However, as mentioned, uniform plating over the width of the wafer may be difficult to obtain.
A number of factors may contribute to non-uniform plating over a face of a substrate. These factors may, for example, include the direction of flow of a plating solution over the face of the substrate, depletion of the plating solution, the positioning of an anode or a cathode which is used for plating, or current density variations over the face of the substrate, particularly current density edge effects around an edge of the substrate. Temperature is also a known factor which influences plating rate. Some of the influences that temperature may have on plating rate are discussed in more detail with respect to FIGS. 5G and 5H in the specification of U.S. patent application Ser. No. 08/452,255 filed May 26, 1995 now abandoned.
Typical problems associated with plating are discussed with reference to FIGS. 1, 2 and 3.
FIG. 1 illustrates a typical plating system 20 which may be used for plating on a lower face 22 of a substrate 24 such as a wafer or a printed circuit board.
The plating system 20 in the illustrated example comprises a tank 26, a substrate holder 28, a pump 30, and an electrical biasing device 32.
The substrate holder 28 locates the substrate 24 over an upper opening of the tank 26 so that the lower surface 22 of the substrate 24 faces downwardly into the tank 26. The tank 26 may be filled with a plating solution which contacts the lower face 22 of the substrate 24.
The electrical biasing device 32 is used for creating a voltage potential between the substrate 24 and the plating solution in the tank 26. By creating a voltage potential between the substrate 24 and the plating solution, a layer is plated on the lower face 22 of the substrate 24.
Due to depletion of the plating solution near the lower face 22 of the substrate 24, it may be necessary to continuously circulate the plating solution through the tank 26. The pump 30 is used to circulate the plating solution through the tank and over the lower face 22 of the substrate. The pump 30 supplies plating solution to a nozzle 34 in the bottom of the tank 26. The nozzle 34 then directs the plating solution upwardly and onto a central region 36 of the lower face 22 of the substrate 24. The plating solution then flows concentrically outwardly over an outer region 38 of the lower face 22 of the substrate 24, and then over an edge region 40 of the lower face 22 of the substrate 24. The plating solution then exists through multiple holes 42 near an edge of the substrate 24, from where the plating solution flows back to the pump 30.
FIG. 2 illustrates a typical profile of a plating layer 44 which is formed on the lower face of the substrate 24 when using a system such as illustrated in FIG. 1. The plating layer 44 is typically relatively thick on the central region 36 of the lower face 22 of the substrate 24. The plating layer 44 then decreases in thickness towards the outer region 38 on the lower face 22 of the substrate 24. The plating layer 44 then typically increases in thickness towards the edge region 40 of the lower face 22 of the substrate 24. The plating layer 44 may vary by 50% or more in thickness, depending on the particular plating system and substrate characteristics. A number of factors contribute to variations in thickness of the plating layer 44.
One factor which may contribute to a non-uniformity in thickness of the plating layer 44 deals with lack of agitation and subsequent replacement of the plating solution after the plating solution has become depleted. The lower face 22 of the substrate 24 is initially exposed to a concentration and chemical composition of the plating solution substantially the same as the plating solution throughout the tank 26. Once plating is initiated the plating solution near the substrate starts losing molecules to the plating layer 44 being formed. The result of the loss of molecules is that a thin boundary layer of depleted plating solution forms on the lower face 22 of the substrate 24. The effect of the boundary layer of depleted plating solution is that less plating occurs as what could be achieved with undepleted plating solution, thus reducing the rate of plating on the lower face 22 of the substrate 24.
A further complication is that the boundary layer does not form uniformly over the lower face 22 of the substrate 24 due to differential agitation of the boundary layer, thus resulting in non-uniform rates of plating over the face of the substrate. Non-uniform formation of the plating layer may be attributed to a number of factors, one of which is differences in agitation of the plating solution over the lower face 22 of the substrate 24.
For example, FIG. 3 illustrates schematically how the plating solution flows concentrically outwardly over the lower face of the substrate 24. The plating solution first flows through an area indicated by the small circle 46A and then through an area indicated by the large circle 46B. When flowing through the small area 46A, the plating solution has a relatively high velocity, indicated by the arrow 48A. However, due to the larger outer area 46B, the plating solution has a velocity, indicated by the arrow 48B, when flowing through the outer area 46B which is less than the velocity 48A when flowing through the inner area 46A. The velocity 48A may be sufficient to cause agitation of the plating solution after having become depleted to such a degree that a depleted boundary layer of the plating solution is continuously replaced with undepleted plating solution. However, the velocity 48B may be of a magnitude which is insufficient for purposes of causing sufficient agitation of the plating solution and may, correspondingly, cause less replacement of a depleted boundary layer of the plating solution. Less plating then, accordingly, occurs as the plating solution flows concentrically outwardly.
A further factor which may contribute to a non-uniformity in thickness of the plating layer 44 may deal with the electrical characteristics of the plating system 20. A higher current density tends to be created near a peripheral edge of the substrate 24. A higher current density, in turn, results in more plating near the peripheral edge of the substrate 24, thus accounting for the plating layer 44 being thicker on the edge region 40 of the lower face 22 of the substrate 24.
Yet a further complication to depletion of the plating solution and the subsequent formation of the boundary layer of depleted plating solution is that may contribute to variations in thickness of the plating layer 44 is depletion of the plating solution while flowing over the lower face 22 of the substrate 24. The plating solution is initially undepleted when flowing over the central region 36, thus accounting for the plating layer 44 being relatively thick on the central region 36. The plating solution then depletes as it flows towards the outer region 38 and the edge region 40, thus accounting for a reduction in thickness of the plating layer 44 on the outer region 38.
The above example illustrates merely some of the electroplating problems which may contribute to a plating layer having non-uniform thickness being formed. Various other factors may contribute to a plating layer being formed having a non-uniform thickness. These factors are not all discussed in detail herein. Suffice it to say that it is an object of the present invention to tailor thickness of plating over a width of a substrate and, more particularly, to ensure more uniform plating over a width of a substrate.